8 Dec
2006
8 Dec
'06
10:43 a.m.
I've read several replies indicating that the drive needs to be taken apart.
People have also advocated just "hitting" the drive to break the stiction.
Holding the drive and giving it a quick twist around the spindle axis has always
worked for me and avoids potential problems with disassembly or damage.
Has anyone seen stiction on the IDE or later drives? The only stiction I've seen
has always been on the 5 1/4" MFM/RLL type drives, and probably ESDI/SCSI/SASI
as well although my experience is limited on those drives.
Something else I've noticed is that if a drive has stiction, that stiction will
return after the drive sets for a while again. Anyone know what actually causes
stiction?
8 Dec
8 Dec
11:07 a.m.
On 8 Dec 2006 at 10:43, Marvin Johnston wrote:
I've read several replies indicating that the
drive needs to be taken apart.
People have also advocated just "hitting" the drive to break the stiction.
I don't think what some of the folks were talking about was
"stiction". It seemed as if the positioner itself were stuck. IIRC.
In the case of stiction, the spindle motor itself cannot overcome the
adhesion between the platters and the heads, so the drive never spins
up.
I've got an SA-4008 that I'm wondering about--obviously the spindle
motor spins the platters up to speed, but the drive doesn't come
ready (it did about 5 years ago when I last powered it). I'm
wondering if there's a stuck positioner issue on that one. Anyone
have any ideas?
Cheers,
Chuck
11:54 a.m.
I've got an SA-4008 that I'm wondering
about--obviously the spindle
You mean the 14" one?
motor spins the platters up to speed, but the drive
doesn't come
ready (it did about 5 years ago when I last powered it). I'm
wondering if there's a stuck positioner issue on that one. Anyone
have any ideas?
The SA4000 that I have a techincal manual for (and which is used in the
PERQ 1) uses a stepper motor to move the heads. It doesn't (AFAIK)
require any feedback signal from the platters to go ready...
Let me look at the manual. I can at least see what causes the ready line
to be asserted, and you can check back from that.
-tony
12:11 p.m.
On 8 Dec 2006 at 19:54, Tony Duell wrote:
Exactly. I'm not sure if 4008 is the right model; it's the 40MB one.
I've got
an SA-4008 that I'm wondering about--obviously the spindle
You mean the 14" one? Let me look at the manual. I can at least see what
causes the ready line
to be asserted, and you can check back from that.
Thanks!
--Chuck
5:19 p.m.
The SA4000 that I have a techincal manual for (and
which is used in the
PERQ 1) uses a stepper motor to move the heads. It doesn't (AFAIK)
require any feedback signal from the platters to go ready...
No, I was wrong. According to the manual, the 'ready' signal depends on a
clock signal that comes from a special track on one of the platters,
picked up by a fixed head. The basic procedure if the drive won't go
ready is to look at the output of this head's amplifier, if that's
missing/incorrect to troubleshoot or replace the read/write PCB (and if
that doesn't get it back, then the fault is inside the HDA and is not
field repairable). If the output of the clock head amplifier is OK, then
therer's a fault on the control PCB.
I can find more details if you want.
-tony
5:44 p.m.
On 9 Dec 2006 at 1:19, Tony Duell wrote:
No, I was wrong. According to the manual, the
'ready' signal depends on a
clock signal that comes from a special track on one of the platters,
picked up by a fixed head. The basic procedure if the drive won't go
ready is to look at the output of this head's amplifier, if that's
missing/incorrect to troubleshoot or replace the read/write PCB (and if
that doesn't get it back, then the fault is inside the HDA and is not
field repairable). If the output of the clock head amplifier is OK, then
therer's a fault on the control PCB.
Thanks, it's enough to get me started. I'll let you know if I'm
unsuccessful.
Cheers,
Chuck
9 Dec
9 Dec
1:07 p.m.
On 9 Dec 2006 at 1:19, Tony Duell wrote:
OK, here are the actual procesdures from the manual :
Check the DC volages at the control PCB J6 connector (to the read/write PCB)
Pin J : +5V
Pin E : -5V
Pin 22 : +24V
Pin B : +12V
(If any of those are misisng/incorrect, there's a procedure to check them
back through the actuator PCB, etc, but I think you can manage that...)
OK, on to the test you need : Drive Not Ready.
It assumes the power voltages are OK, and that the disk is rotating.
Notes :
1) Ready indication will appea as a logical 1 on TP28 of the control PCB.
2) Ready will not become active for 1.5 minutes after AC and DC power are
applied
3) The drive must be selected befroe Ready is available at the interface
connector
Test 1 : Check clock track amplitude at control PCB TP9 (must be more
than 1.5V)
Test 1 OK?
No : Replace Read/Wrtie PCB [1]
Test 1 OK ?
No : Problem with clock head -- not field repairable
Yes : go to test 2
Yes : Got to test 2
Test 2 : Check TP5 on control PCB for TTL level 500ns puleses every 2.2us
Test 2 OK ?
No : Replace control PCB [2]
Yes : go to test 3
Test 3 : Check TP27 on control PCB for TTL level square wave of 70ns
+/-5ns (14.2MHz)
Test 3 OK ?
No : Replace control PCB [3]
Yes : go to test 4
Test 4 : Check TP26 of control PCB for a TTL level pulse 1.1us duration
every 2oms (Index)
Test 4 OK ?
No : Replace control PCB [4]
Yes : Test done
ARD#'s notes
[1] Check circuitry round 6B and 6D on the read/write PCB, these being
the clock head amplifiers
[2] Check round 2A and 1B(section b) on the control PCB.
[3] This is the output of the phase-lock oscillator. Check Q3, 3D
(oscillator), 2B, Q1, Q2 (phase detector), etc
[4] Check counter 2E, register 1G, and associated gates
-tony
No, I was wrong. According to the manual, the
'ready' signal depends on a
clock signal that comes from a special track on one of the platters,
picked up by a fixed head. The basic procedure if the drive won't go
ready is to look at the output of this head's amplifier, if that's
missing/incorrect to troubleshoot or replace the read/write PCB (and if
that doesn't get it back, then the fault is inside the HDA and is not
field repairable). If the output of the clock head amplifier is OK, then
therer's a fault on the control PCB.
Thanks, it's enough to get me started. I'll let you know if I'm
unsuccessful.
10 Dec
10 Dec
10:59 a.m.
On 9 Dec 2006 at 21:07, Tony Duell wrote:
If you need any more help with the electronics of the drive itself, I
have the service manual. It's got some board-swapping flowcharts in the
first part (that's what I produced that 'procedure' from), but the second
bit is the schematics. So I can probably figure it out.
But if the problem is inside the HDA (maybe a damaged clock head or
track, this AFAIK is a fixed head that always reads the same track off
one of the platters) or if the problem is in the cotnroller, which I know
nothing about (what is the machine, BTW?) then I probably won't be much use.
-tony
OK, here are the actual procesdures from the
manual : [snip]
Many thanks--I've got your notes safely stashed away. Hopefully it's
not the custom controller in the same box...
1:42 p.m.
On 10 Dec 2006 at 18:59, Tony Duell wrote:
But if the problem is inside the HDA (maybe a damaged
clock head or
track, this AFAIK is a fixed head that always reads the same track off
one of the platters) or if the problem is in the cotnroller, which I know
nothing about (what is the machine, BTW?) then I probably won't be much use.
It's an external disk drive for a business micro. Protocol is GPIB
and the controller is a very large PCB with lots of bit-slice logic
on it, bearing the logo "Microsystems Associates".
Cheers,
Chuck
8 Dec
8 Dec
11:09 a.m.
Stiction is mostly caused by a breakdown in the lubricant.
Over simplified, oil turns to varnish or tar.
I too have had good luck with a wrist twist to get them spinning again.
and there is a very high likeliness that they will bind up again if allowed to stop and
cool down.
I have seen several IDE's have had stiction problems after being shut down for the
first time after extended 24/7 use.
I have a maxtor that bound up on me, two days ago, after 6 years of constant 24/7 server
use I shut it down for a long over due
upgrade. It cooled off while I moved the system to the my cave to work on it and the drive
never spun back up.
I knew I should have backed a few things up before I moved it :(
In the past, I opened up an IDE and spun it up by hand. Once it got running, I got the
files off of it I needed.
For fun I left it run open on the workbench bench for close to a week before it died.
I had a lot of people come by to see it, and stared in disbelief as the heads sang across
the disk when I did a defrag on it.
later
Bob Bradlee
On Fri, 08 Dec 2006 10:43:08 -0800, Marvin Johnston wrote:
I've read several replies indicating that the drive
needs to be taken apart.
People have also advocated just "hitting" the drive to break the stiction.
Holding the drive and giving it a quick twist around
the spindle axis has always
worked for me and avoids potential problems with disassembly or damage.
Has anyone seen stiction on the IDE or later drives?
The only stiction I've seen
has always been on the 5 1/4" MFM/RLL type drives, and probably ESDI/SCSI/SASI
as well although my experience is limited on those drives.
Something else I've noticed is that if a drive has
stiction, that stiction will
return after the drive sets for a while again. Anyone know what actually causes
stiction?
9 Dec
9 Dec
7:11 a.m.
From: "Bob Bradlee" <caveguy at
sbcglobal.net>
Stiction is mostly caused by a breakdown in the lubricant.
Over simplified, oil turns to varnish or tar.
Hi
I have several friends that worked at Seagate when they had
problems of stiction. It was not a lubricant problem. It was
caused by the surfaces being too smooth. When to really
smooth surfaces sit together for a long time, the air is squeezed
out. Once the surfaces really touch, there is a thing called
molecular adhesion.
Anyone that has worked with guage blocks is familair with
this.
Seagate fixed the problem by roughing the surfaces enough
so that they didn't quite sqeeze out enough air to adhere.
The surfaces are lubed but not with petrolium greases
so they don't break down to tars. As I recall, they used some
type of synthetic oil and only in tiny amounts. It was just
to keep the surfaces apart while it spun up. After that,
the head was flying and no longer made contact.
Dwight
_________________________________________________________________
All-in-one security and maintenance for your PC.? Get a free 90-day trial!
http://clk.atdmt.com/MSN/go/msnnkwlo0050000002msn/direct/01/?href=http://cl…
8:13 a.m.
The problems i saw were with old drives that ran 24/7 and were shut down to be moved to a
new location.
The problem occured when they cooled down. Once warmed back up, they could be powered down
and back up
as long as they did not cool down again.
These were systems that had 30k -> 50k+ hours on them in 24/7 operation.
I always figured it was bearings or bearing related lub problems. I have lost track of the
number of cheep power
supply and cpu fans that required a spin to get start again after they had cooled down.
In the late 90's the cheep cpu fans they were putting in clones would last between 15k
-> 20k hours. Changing
fans kept some small service shops in business for several years around the turn of the
century.
I had never thought of the platters and heads being a cause of stiction.
In a crude attempt to put things back on topic :-)
My friend David had a stiction problem last year on a 3344 disk pack running on an IBM
System 3. The head
retracted to park and stuck to a very old and very soft rubber home stop. It did not
release and come ready untill
pushed from behind with a stick, via the inspection hole, to get it moveing again.
Stiction comes in many forms from numerious reasons .....
Just some random thoughts
Bob Bradlee
On Sat, 09 Dec 2006 07:11:25 -0800, dwight elvey wrote:
>From: "Bob Bradlee" <caveguy at
sbcglobal.net>
>
>Stiction is mostly caused by a breakdown in the lubricant.
>
>Over simplified, oil turns to varnish or tar.
>
Hi
I have several friends that worked at Seagate when they had
problems of stiction. It was not a lubricant problem. It was
caused by the surfaces being too smooth. When to really
smooth surfaces sit together for a long time, the air is squeezed
out. Once the surfaces really touch, there is a thing called
molecular adhesion.
Anyone that has worked with guage blocks is familair with
this.
Seagate fixed the problem by roughing the surfaces enough
so that they didn't quite sqeeze out enough air to adhere.
The surfaces are lubed but not with petrolium greases
so they don't break down to tars. As I recall, they used some
type of synthetic oil and only in tiny amounts. It was just
to keep the surfaces apart while it spun up. After that,
the head was flying and no longer made contact.
Dwight
_________________________________________________________________
All-in-one security and maintenance for your PC. Get a free 90-day trial!
http://clk.atdmt.com/MSN/go/msnnkwlo0050000002msn/direct/01/?
href=http://clk.atdmt.com/MSN/go/msnnkwlo0050000001msn/direct/01/?href=http…
sc_cid=msn_hotmail
6:03 p.m.
In the late 90's the cheep cpu fans they were
putting in clones would last between 15k -> 20k hours. Changing
fans kept some small service shops in business for several years around the turn of the
century.
Same here back then from early 90's to 2000 or so, we replace fans on
anything from 40mm all the way to 80mm bec of bearings dying. Now
no longer since we just replace the heatsink or a power supply.
My friend David had a stiction problem last year on a
3344 disk pack running on an IBM System 3. The head
retracted to park and stuck to a very old and very soft rubber home stop. It did not
release and come ready untill
pushed from behind with a stick, via the inspection hole, to get it moveing again.
Stiction comes in many forms from numerious reasons .....
Bob Bradlee
I had several micropolis 13xx series get rubber stop stiction that
glued the arm to that rubber stop and not letting arm go. Had to
open it and push arm off rubber for data recovery.
Cheers, Wizard
9:18 a.m.
On 9 Dec 2006 at 7:11, dwight elvey wrote:
I have several friends that worked at Seagate when
they had
problems of stiction. It was not a lubricant problem. It was
caused by the surfaces being too smooth. When to really
smooth surfaces sit together for a long time, the air is squeezed
out. Once the surfaces really touch, there is a thing called
molecular adhesion.
Anyone that has worked with guage blocks is familair with
this.
That's the story that I got from the Seagate marketing engineer when
I complained about new ST-225's occasionally showing this problem. .
However, Wikipedia states that the problem really is heat and
lubricants:
http://en.wikipedia.org/wiki/Stiction
So, the moral is "never trust a marketing guy", I guess.
Cheers,
Chuck
8 Dec
8 Dec
11:26 a.m.
Marvin Johnston wrote:
Has anyone seen stiction on the IDE or later drives?
Yep, on early Connor IDE drives. I don't know if that happened to be a bad
batch (I was testing them a couple of years after manufacture and 2/3 were DOA
due to stiction), or "they were all like that, sir".
I don't recall seeing it on later IDE drives though, or on any SCSI drive
(don't forget that some SCSI drives won't spin up without communication with
the HBA, so can appear dead when power only is applied)
Something else I've noticed is that if a drive has
stiction, that stiction will
return after the drive sets for a while again. Anyone know what actually causes
stiction?
my *assumption* is that it's some sort of break-down in either the head
material or the coating on the platters, such that when at rest of any period
of time the heads manage to reattach themselves to the platters.
Unfortunately there seems to be no cure, expect for never powering the drive
down until some other failure finally renders it useless.
cheers
Jules
--
there's a carp in the tub
there's a carp in the tub
so nobody's taken a bath
1:52 p.m.
Marvin Johnston wrote:
<snip>
Has anyone seen stiction on the IDE or later drives? The only stiction I've seen
has always been on the 5 1/4" MFM/RLL type drives, and probably ESDI/SCSI/SASI
as well although my experience is limited on those drives.
I had three cases out of it on ST-225 and relatives.
Something else I've noticed is that if a drive has
stiction, that stiction will
return after the drive sets for a while again. Anyone know what actually causes
stiction?
In the case of the ST-225 they used defective magnetic lubricant on the
axis which eventually
developed too much friction for the start up pulses in the motor to
overcome. A bit of a kick
as you mentioned is the only thing to try. once the disk stack is in
motion, the motors have to
be able to keep it going and if the friction is that great there is no
way that taking it apart will
fix anything.
Never saw it in the smaller drives, but I did see a number of them with
being unable to regulate their
speed, which may have been a vaiant. They would make different sounds
after startup and
eventually spin down when they could not find a place to track.
This could have been a failure to spin at the right speed, or a failure
to correctly lock
onto the reference servo and see a basic clock, or an increased friction
keeping it from
getting up to the correct speed.
Jim
5:59 p.m.
Marvin Johnston wrote:
Something else I've noticed is that if a drive has stiction, that stiction
will return after the drive sets for a while again. Anyone know what actually
causes stiction?
my dim recollection (back in the day I did actually talk to people from
rodime, lapine, nec, etc about 10 & 20mb 3" st-506 drives) is that it
was combination of the media plating, coating on the platter and the
heads themselves.
since there is no landing zone the heads set down on the media. if the
media was very smooth and the lubricant too thick, the heads would glue
themselves to the media.
I suspect there is slightly more to it than that, but it's been a while
for me. I know the firmware often did a lot to try and mitigate the
problem but I think in the end it was changes in lubricant/coating which
made it go away.
-brad
9 Dec
9 Dec
8:40 a.m.
On Dec 8, 2006, at 8:59 PM, Brad Parker wrote:
Ahh yes that corresponds exactly with my memory of the situation
as well.
-Dave
--
Dave McGuire
Cape Coral, FL
Something else
I've noticed is that if a drive has stiction, that
stiction
will return after the drive sets for a while again. Anyone know
what actually
causes stiction?
my dim recollection (back in the day I did actually talk to people
from
rodime, lapine, nec, etc about 10 & 20mb 3" st-506 drives) is that it
was combination of the media plating, coating on the platter and the
heads themselves.
since there is no landing zone the heads set down on the media. if
the
media was very smooth and the lubricant too thick, the heads would
glue
themselves to the media.
I suspect there is slightly more to it than that, but it's been a
while
for me. I know the firmware often did a lot to try and mitigate the
problem but I think in the end it was changes in lubricant/coating
which
made it go away.
8:39 a.m.
On Dec 8, 2006, at 1:43 PM, Marvin Johnston wrote:
I've read several replies indicating that the
drive needs to be
taken apart.
People have also advocated just "hitting" the drive to break the
stiction.
Holding the drive and giving it a quick twist around the spindle
axis has always
worked for me and avoids potential problems with disassembly or
damage.
Has anyone seen stiction on the IDE or later drives? The only
stiction I've seen
has always been on the 5 1/4" MFM/RLL type drives, and probably
ESDI/SCSI/SASI
as well although my experience is limited on those drives.
Something else I've noticed is that if a drive has stiction, that
stiction will
return after the drive sets for a while again. Anyone know what
actually causes
stiction?
The story I heard "back in the day" (which was at work, from a
service bulletin of some sort, so I treat it with some credibility)
is that designers chose the spindle lubricant unwisely in some models
of drives, and it spun out onto the platters a bit and gummed up.
-Dave
--
Dave McGuire
Cape Coral, FL
10 Dec
10 Dec
7:36 a.m.
From: Dave McGuire <mcguire at neurotica.com>
On Dec 8, 2006, at 1:43 PM, Marvin Johnston wrote:
Hi
I find this hard to believe. If any of the lub leaked onto the
surface, it would surely cause the head to crash one spin
up. On the 225 I had, it always worked fine after a smack
on the side. Anything like steaks of lub from the bearing
would have destroyed the disk in no time.
Dwight
_________________________________________________________________
Get free, personalized commercial-free online radio with MSN Radio powered
by Pandora http://radio.msn.com/?icid=T002MSN03A07001
I've read several replies indicating that the
drive needs to be taken
apart.
People have also advocated just "hitting" the drive to break the
stiction.
Holding the drive and giving it a quick twist around the spindle axis has
always
worked for me and avoids potential problems with disassembly or damage.
Has anyone seen stiction on the IDE or later drives? The only stiction
I've seen
has always been on the 5 1/4" MFM/RLL type drives, and probably
ESDI/SCSI/SASI
as well although my experience is limited on those drives.
Something else I've noticed is that if a drive has stiction, that
stiction will
return after the drive sets for a while again. Anyone know what actually
causes
stiction?
The story I heard "back in the day" (which was at work, from a service
bulletin of some sort, so I treat it with some credibility) is that
designers chose the spindle lubricant unwisely in some models of drives,
and it spun out onto the platters a bit and gummed up.
-Dave
10:36 a.m.
On Dec 10, 2006, at 10:36 AM, dwight elvey wrote:
I find this hard to believe. If any of the lub leaked onto the
surface, it would surely cause the head to crash one spin
up. On the 225 I had, it always worked fine after a smack
on the side. Anything like steaks of lub from the bearing
would have destroyed the disk in no time.
I'd have thought that too, but that's what the bulletin said so I
believed it at the time. I don't know what the flying heights were
in those old drives but I'm sure they were nowhere near as low as
modern drives.
But you mentioned "225"...As in Seagate ST-225? I've never heard
of stiction problems with those (I have used quite a few of them, and
I currently have a few in DEC machines as RD-31s)...I wasn't aware
that they could develop such problems. I will keep an eye on mine!
-Dave
--
Dave McGuire
Cape Coral, FL
I've read several replies indicating that the
drive needs to be
taken apart.
People have also advocated just "hitting" the drive to break the
stiction.
Holding the drive and giving it a quick twist around the spindle
axis has always
worked for me and avoids potential problems with disassembly or
damage.
Has anyone seen stiction on the IDE or later drives? The only
stiction I've seen
has always been on the 5 1/4" MFM/RLL type drives, and probably
ESDI/SCSI/SASI
as well although my experience is limited on those drives.
Something else I've noticed is that if a drive has stiction,
that stiction will
return after the drive sets for a while again. Anyone know what
actually causes
stiction?
The story I heard "back in the day" (which was at work, from a
service bulletin of some sort, so I treat it with some
credibility) is that designers chose the spindle lubricant
unwisely in some models of drives, and it spun out onto the
platters a bit and gummed up.
16 Dec
16 Dec
11:53 a.m.
From: Dave McGuire <mcguire at neurotica.com>
---snip
The story I heard "back in the day" (which was at work, from a
service bulletin of some sort, so I treat it with some credibility) is
that designers chose the spindle lubricant unwisely in some models of
drives, and it spun out onto the platters a bit and gummed up.
I find this hard to believe. If any of the lub leaked onto the
surface, it would surely cause the head to crash one spin
up. On the 225 I had, it always worked fine after a smack
on the side. Anything like steaks of lub from the bearing
would have destroyed the disk in no time.
-Dave
--
Dave McGuire
Cape Coral, FL
_________________________________________________________________
MSN Shopping has everything on your holiday list. Get expert picks by style,
age, and price. Try it!
http://shopping.msn.com/content/shp/?ctId=8000,ptnrid=176,ptnrdata=200601&a…
6582
days inactive
6590
days old
22 comments
10 participants
participants (10)
-
ard＠p850ug1.demon.co.uk -
brad＠heeltoe.com -
caveguy＠sbcglobal.net -
cclist＠sydex.com -
dkelvey＠hotmail.com -
jpero＠sympatico.ca -
julesrichardsonuk＠yahoo.co.uk -
jwstephens＠msm.umr.edu -
marvin＠rain.org -
mcguire＠neurotica.com